# Falcon

An on-device speaker diarization engine from Picovoice that identifies distinct speakers and their speaking intervals in an audio recording, with SDKs for multiple languages and platforms.

Falcon is an on-device speaker diarization engine: given an audio recording, it identifies distinct speakers and divides the audio into segments marked with which speaker was talking and when. All processing runs locally, with no audio sent to a server.

It ships as SDKs for Python, C, Web, iOS and Android, and runs on Linux (x86_64), macOS (x86_64 and arm64), Windows (x86_64 and arm64), Raspberry Pi (3, 4, 5), Android, iOS, and in-browser via Chrome, Safari, Firefox and Edge. Demos are provided for each supported platform.

Using it requires a Picovoice AccessKey, which needs an internet connection to validate and to enforce account usage limits, even though the diarization itself runs entirely offline.

Website: https://github.com/Picovoice/falcon